#hazzy Logs

Mar 28 2020

03:11 AM TurBoss: morning
03:11 AM * TurBoss uploaded an image: Captura de pantalla de 2020-03-28 09-10-55.png (195KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/FWTnsiyJqOcxXPwZrPeSRcbI >
03:16 AM Not-8b5f: [02qtpyvcp] 07TurBoss pushed 031 commit to 03VCP_API [+0/-0/±1] 13https://git.io/JvQIO
03:16 AM Not-8b5f: [02qtpyvcp] 07TurBoss 03a42d179 - iterate over folders 1 level
07:19 AM hazzy-m: TurBoss nice!
07:20 AM TurBoss: this stuff is interesting
07:20 AM TurBoss: told to a friend
07:56 AM TurBoss: ```
07:56 AM TurBoss: pip3 install opcua-modeler
08:12 AM Lcvette: morning o/
08:12 AM Lcvette: what is it?
08:12 AM Lcvette: :o
08:34 AM TurBoss: morning Lcvette
08:35 AM TurBoss: comunication system
08:40 AM jthornton: morning
08:41 AM TurBoss: hello JT
09:21 AM Guido[m]: Morning, i was wondering soth regarding PB and ATC: Can it deal with toolnumber != pocket number (not equal) - i am looking for approches to get the tool number of the pocket of the tool currently in the spindle and #current_pocket is only giving the row number in stock master
09:31 AM TurBoss: hello, as far I know the pocket of the tool in spindle goes to 0
09:32 AM TurBoss: the atc subroutines store tool pockets in the var file
09:32 AM TurBoss: they are not the same in the tool table
09:33 AM TurBoss: Lcvette may know better
09:47 AM * TurBoss uploaded an image: Captura de pantalla de 2020-03-28 15-46-34.png (120KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/tqKbtVOhpUNBPepoFUcLymFp >
09:49 AM Lcvette: Guido: the atc currently ignores the tool tables pocket number
09:49 AM Lcvette: it manages pocket numbers based on the var file parameter numbers listed in the toolchange.ngc subroutine
09:50 AM Lcvette: Hazzy is working on a new atc that may be different though i think
09:50 AM Guido[m]: okay, thanks - guess I have to continue finding work aorunds for that topic... (working on a rack style with remap)
09:50 AM Lcvette: woul be the same
09:50 AM Guido[m]: out of curiosity: Is it possible to store data from a macro in a global variable that is somewhat persistent (i.e. the next M6 remap call can access that data?)
09:50 AM Lcvette: just a different method to go to the selected tool
09:51 AM Lcvette: thats what the toolchange.ngc does
09:51 AM Lcvette: it is storing the tool numbers in the var file in persistent locations
09:52 AM Lcvette: 5190-5201 in the sim atc remappe subroutine toolchange.ngc those parameters are persistent and manage the tools
09:53 AM Lcvette: no need to do anything different except manage the code on how the machine moves to the specified tools pocket instead of rotating the atc carousel to arrive at the correct tool
09:54 AM Lcvette: Guido: ^^^
10:44 AM Not-8b5f: [02qtpyvcp] 07TurBoss pushed 031 commit to 03VCP_API [+0/-0/±1] 13https://git.io/JvQnQ
10:44 AM Not-8b5f: [02qtpyvcp] 07TurBoss 034c73a92 - float valriables
01:51 PM Guido[m]: @Lcvette: Sorry, I got sidelined. Thank you. Ill have a look at that
04:21 PM ewildgoose[m]: <Lcvette "and again G59.3 MUST always be x"> Ok. Thanks this was the trouble.
04:23 PM ewildgoose[m]: I'm asking before I try and edit myself but: is there a reason the probe routine doesn't function without leaving the current offset space? Just eyeballing it seems possible to stay in G54 or whatever? Should I have a try and submit a change? Bigger picture is there a reason it can't be done?
04:34 PM * ewildgoose[m] sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/rkrJJsyEHXibdsBRKMdSoLyK >
04:37 PM ewildgoose[m]: What was the question? Ooh. So this type of probe doesn't seem to fit into the current probe workflow? How would you recommend to add it?
04:37 PM ewildgoose[m]: Do you think a new screen works better? Could have the parameters for size of the block, etc? Or do you have some inspiration on using within the current probing framework?
04:37 PM ewildgoose[m]: Ta
04:40 PM ewildgoose[m]: PS. Used my "real probe" yesterday for some jobs (ie not the thing above, but something you put in the spindle). Needed to center find on some circular tube. In the past I was doing manual touch offs in each dimension. Was soooooooo much easier using your probe routines!! 👍👍👍😁
04:43 PM ewildgoose[m]: Right now it's a bit of a pain to use because it's logic is inverted compared with the tool length sensor, so I need to edit the hal file to swap it in. Tomorrow's job is to build and electrical inverter for one of the switch inputs... Thinking PNP transistor circuit but I'm no electrical wizz... Suggestions welcome!
05:02 PM Lcvette: @ewildgoose:matrix.org: you can change whatever you need to for yourself, but it works flawlessly otherwise
05:03 PM Lcvette: it has to change into the G59.3 in order for the recorded distance to be in machine coordinates
05:04 PM Lcvette: this keeps all of the tool lengths standard to the spindle nose height
05:05 PM Lcvette: if your machine uses a er type collet with non repeatable holder you may be better changing to suit your own needs but for most with repeatable length tool holders the way it is setup is the correct way so that all tools measure against a known height
05:06 PM Lcvette: it also uses this method to use the probe tool as a master tool so that when using a touch probe for finding part zero locations ALL tools are properly set with minimal input
05:06 PM Lcvette: jope that helps
06:24 PM hazzy-m: Guido: https://github.com/kcjengr/qtpyvcp/tree/ATC_Remap
06:25 PM hazzy-m: In that branch there’s an example of how you can load external data into the interpreter
06:25 PM hazzy-m: They read the tool table file and stores that information in the dictionary so you can have the actual tool number in pocket number available as interpreter parameters
06:26 PM hazzy-m: Sorry, using voice to text, I’ll be at dev in a bit
06:39 PM ewildgoose[m]: <Lcvette "it also uses this method to use "> Understood. But this type of probe has a slightly different workflow. Here the tool stays put (regardless of tool), and you measure Z offset with your probe (you might move your tool setter to the top of the work say). However, it needs compensation for the height of the probe in this case
06:46 PM Lcvette: @ewildgoose:matrix.org: sure, just modify how you need it. You could add a line in the subroutine to account for the tool plate height
06:48 PM ewildgoose[m]: <Lcvette "@ewildgoose:matrix.org: you can "> I think if you try the following you may find a "bug" (although I need to double check now):
06:49 PM ewildgoose[m]: I believe I can rewrite this to use absolute coordinates... I will try and submit a pull
06:49 PM Lcvette: If you can that would be great
06:50 PM Lcvette: I could not
06:50 PM ewildgoose[m]: <Lcvette "@ewildgoose:matrix.org: sure, ju"> I'm not sure I follow your suggestion? I don't see this works of you have both a probe AND touch plate? You need to distinguish which you use no?
06:51 PM Lcvette: Probe input is the same for both
06:52 PM ewildgoose[m]: I'm thinking that the touch plate is kind of a separate probe type really. Perhaps a whole new screen, similar to the current probe interface but with the dimensions of your touch plate? Keeps it separate to the current probe interface?
06:52 PM ewildgoose[m]: Just thought you might have a more elegant integration suggestion?
06:53 PM Lcvette: They're is already a touch plate
06:53 PM Lcvette: It's on the tool page
06:53 PM ewildgoose[m]: <Lcvette "Probe input is the same for both"> Nope you have lost me now? Above you say edit the sub routine? I presumed you meant to hard code the touch plate thickness? How are you suggesting to distinguish when one uses touch plate and when one uses probe?
06:54 PM Lcvette: And inputs for it are on the offset page
06:54 PM ewildgoose[m]: <Lcvette "They're is already a touch plate"> Aha! Ok. Interesting? What tool is yours? How do you see that working though? I'm a bit lost?
06:54 PM Lcvette: That's for measuring tool lengths
06:55 PM Lcvette: Probe is for a probe tool
06:55 PM ewildgoose[m]: Am I perhaps not on the same version as you again??
06:55 PM Lcvette: What version are you on?
06:56 PM Lcvette: It's on the latest stable
06:56 PM ewildgoose[m]: About 24 hours old master?
06:56 PM Lcvette: Probe page is for a touch probe
06:56 PM Lcvette: Tool page and posts page is for a tool setter
06:57 PM * ewildgoose[m] sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/kgqTKSQlRsOitQUiGBCVDuwB >
06:57 PM ewildgoose[m]: I only see that we have options for 2 of these at present? I think there is a case for all three. At least for machines without tool holders?
06:59 PM ewildgoose[m]: I'll have a crack at altering the probe routine to use only absolute coords. Will feedback results
07:02 PM Lcvette: Yeah I'm not sure about 3 different things
07:02 PM Lcvette: Seems you only need 2 of the three
07:03 PM Lcvette: If you have a probe tool, and a tool setter it seems it should do everything you need
07:03 PM ewildgoose[m]: I also have on my to-do list: rewriting the Linuxcnc VFD stuff... Why do we have so many near identical modules?? Sigh... Also none support my cheapo Nowforever VFD...
07:04 PM Lcvette: If you aren't using the probe tool, then change it for the touch plate
07:04 PM ewildgoose[m]: <Lcvette "Seems you only need 2 of the thr"> Only if you have very precise tool lengths... Sadly...
07:04 PM Lcvette: Or add a subcallbbutton and make another subroutine for the z part top touchplate thing
07:05 PM ewildgoose[m]: I think it's more common in mills to have such spindles. For routers and perhaps plasma cutters, it's more common to touch off the tool to the material instead
07:06 PM ewildgoose[m]: Yeah. I'm thinking a whole extra tab. Similar to the probe interfaces yet with completely different config.
07:07 PM ewildgoose[m]: Could share the actual probe gcode though since ultimately it's just about different offsets, not different probing
07:08 PM ewildgoose[m]: Ok. Thanks. I know enough for now. Will try some more experiments here.
07:08 PM Lcvette: If you add a button for it to touch off on the plate and calculate the plate thickness in the sub i think that's what you ate looking for
07:13 PM ewildgoose[m]: Quick question. On the right side there is a tabbed box. Can be changed to jog, etc. There are two buttons at the bottom... Forgotten their names, something like touch and pulse? What do they do?
07:16 PM ewildgoose[m]: <Lcvette "If you add a button for it to to"> Actually another nice thing you can do with certain touch plates is also corner finding. There is then an offset for probe width plus offset for tool. Faster if tool changing is slow...
07:26 PM Lcvette: Those buttons are for testing
07:26 PM Lcvette: They trigger the probe input in the sim
07:27 PM Lcvette: Use touch
08:02 PM -!- #hazzy mode set to +o by ChanServ
08:06 PM -!- #hazzy mode set to +o by ChanServ
10:30 PM Not-8b5f: [02probe_basic] 07adamj12b forked the repository - 13https://git.io/JvQXB